The thalassemias and hemoglobinopathies represent a heterogeneous group of anemias characterized by absent/reduced or abnormal production of one or more of the globin-molecule subunits, respectively, and strategies which aim to replace the absent or defective globin gene have long been envisioned as potentially curative. Indeed, retroviral vectors carrying globin genes were among the first gene transfer vectors to be tested in murine models, but low gene transfer rates and poor globin gene expression plagued the field. Furthermore, rodent models proved insufficient to model human hematopoieisis. In large animals, significant advances in gene transfer technology have been made by systematically testing transduction methods in a competitive repopulation model, with long-term in-vivo gene transfer levels of 5-10% or higher now achievable after ablative condition with high dose irradiation. The finding of common integration sites among myeloid and erythroid colonies as well as peripheral blood T and B cell populations along with the prolonged contribution of some clones to myeloid progeny satisfied strict criteria for transduction of true hematopoietic stem cells, and the clonal dynamics supported a stochastic model of in vivo hematopoiesis. The development of techniques for clonal tracking have also proven important in assessing the risk of insertional mutagenesis with integrating retroviral vectors. Concurrent with this progress, the Sadelain group succeeded in attaining high titer, stable viral vectors which faithfully deliver the human beta-globin gene along with key regulatory elements sufficient to ameliorate disease in a murine model of beta-thalassemia8, setting the stage for preclinical testing in the large animal model. In collaboration with the Sadelain laboratory, we have now moved forward with preclinical testing of lentiviral gene transfer vectors carrying human beta-globin along with key regulatory sequences in the rhesus macaque model. A number of other issues, however, remain to be addressed prior to clinical application. We have recently established steady state marrow, the only practical stem cell source in sickle cell disease, as a viable target for genetic manipulation in the nonhuman primate9, yet the type and degree of conditioning required to achieve adequate engraftment remains to be established. We have previously shown that low dose irradiation is sufficient to allow clinically relevant levels of engraftment of genetically modified cells in the murine model, even when xenogeneic genes are expressed. Such irradiation doses allowed for long term engraftment by genetically modified cells in the nonhuman primate, but at levels too low to expect clinical benefit. Increasing the irradiation dose to levels bordering myeloablative resulted in only modest improvement. Busulfan is an alkylating chemotherapeutic agent that has long been used as an alternative agent to total body irradiation for conditioning for bone marrow transplantation. However, erratic absorption of the oral formulation necessitated close pharmacokinetic monitoring of individual patients to achieve predictable myelosuppression. We have recently evaluated a newly available intravenous formulation of busulfan in the murine model, and the results demonstrate that dose dependent engraftment can be achieved at levels of up to 80% at nonmyeloative doses. Further improvement can be achieved by delaying infusion to the day of the neutrophil nadir. This agent is now being tested in the nonhuman primate model in an attempt determine the dosage adequate to allow engraftment of genetically modified cells at levels sufficient for clinical application. Three rhesus macaques have recently been transplanted with autologous peripheral blood stem cells transduced with a lentiviral vector carrying the beta globin gene and key regulatory elements. Though initial engraftment was robust, long term levels of genetically modified cells are below that necessary for phenotypic correction and additional measures to produce vectors for this application are now underway. These current studies will be used in support of eventual clinical studies in globin disorders with the ultimate goal of providing preclinical safety and efficacy data in order to maximize the likelihood of success in the context of an acceptable risk/benefit ratio.